Clinical Epidemiology of T-cell Acute Lymphoblastic Leukaemia

Abstract

T-cell acute lymphoblastic leukaemia (T-cell ALL) is a rare and aggressive disease in children and adults. Although overall survival rates reach ~85% with contemporary treatment and stratification, relapse occurs in ~20% of cases, and these patients still have poor survival. Demographic, clinical, and genetic risk and prognostic factors have been proposed, but they are rarely validated. Minimal/measurable residual disease (MRD) is widely used for risk stratification; however, it is not helpful with relapsed patients and has limited accuracy. Therefore, there is a need to identify additional prognostic factors in T-cell ALL. This thesis aimed to investigate clinical factors (sex, age, white blood cell count at diagnosis, central nervous system involvement, organomegalies, early marrow response, MRD response, and genetics) in paediatric/young adult UKALL trials (UKALL VIII to UKALL 2011) for the improvement of current MRD stratification. Treatment response variables (early marrow response and MRD response) were the only significant prognostic factors among other variables in T-cell ALL. D8BM% model, containing only bone marrow percentage measured at day 8, was effective in separating risk groups at very early stage of treatment in UKALL 97/99, UKALL 2003, and UKALL 2011. More importantly, this model improved MRD prediction for clinical events. Dynamic prognostic models with Nonlinear Mixed-Effects (NLME) and Area Under the Curve (AUC) methods were successfully applied to data with treatment response variables at multiple time points; however, they did not have a superior advantage over single-time point measurements. Additionally, Tcell ALL subtypes detected by laboratory techniques and unsupervised clustering algorithm did not have prognostic importance. In summary, treatment response variables are crucial to stratifying T-cell ALL patients other than genetic abnormalities/subtypes. Early treatment response adds prognostic value to MRD stratification at the end of induction, so it can be efficient to stratify T-cell ALL patients at very early stage of treatment.